tbitai/bayes-enron1-spam

Bayes Enron1 spam

This model stores the Bayesian spam probabilites calculated with Graham's algorithm on the Enron1 dataset, as described in the Bayes or Spam? project.

Data structure

probs.json is dumped from a dictionary contaning token: probability pairs, e.g. "enron": 0.0.

The unknown token is [UNK].